OUP user menu

Effect of metoprolol and ivabradine on left ventricular remodelling and Ca2+ handling in the post-infarction rat heart

Michał Mączewski, Urszula Mackiewicz
DOI: http://dx.doi.org/10.1093/cvr/cvn057 42-51 First published online: 1 March 2008


Aims β-Blockers reduce mortality and morbidity in heart failure. Many of their benefits can be explained solely by heart rate reduction (HRR). We aimed to verify whether the β-blocker, metoprolol, and the pure heart-rate-reducing agent, ivabradine, have the same effects on haemodynamic function, ventricular remodeling, and Ca2+ handling in post-myocardial infarction (MI) heart failure in rat.

Methods and results Metoprolol (250 mg/kg/day) or ivabradine (10 mg/kg/day), offering similar HRR, or no treatment, was started 24 h after an induction of MI or sham surgery in rat. Eight weeks post-MI metoprolol and ivabradine similarly partially prevented deterioration of left ventricular (LV) ejection fraction and reduced post-MI LV wall stress. However, metoprolol partially prevented LV dilation, whereas ivabradine potentiated LV hypertrophy. Metoprolol, but not ivabradine, partially prevented post-MI chronotropic incompetence. Metoprolol markedly, whereas ivabradine mildly, increased the amplitude of the Ca2+ transient in post-MI cardiomyocytes. Ivabradine, but not metoprolol, partially prevented the MI-induced depression of sarcoplasmic reticulum Ca2+-ATPase (SERCA) activity, while metoprolol, but not ivabradine, suppressed Na+/Ca2+ exchanger (NCX) overactivity and normalized Ca2+ sensitivity of ryanodine receptors.

Conclusion Although both metoprolol and ivabradine comparably prevented post-MI deterioration of haemodynamic function in the rat, metoprolol had additional potentially beneficial effects; it prevented LV dilation and hypertrophy, chronotropic incompetence, strongly increased contractility of isolated cardiomyocytes, and prevented the potentially proarrhythmic increase in NCX activity. This indicates that pure HRR does not account for effects of β-blockade in the post-MI setting. Metoprolol and ivabradine similarly improve LV function, although differently affect LV morphology and cellular Ca2+ handling in the post-infarction rat heart.

  • Myocardial infarction
  • Remodelling
  • Calcium handling
  • Haemodynamics
  • Heart failure
  • Heart rate

1. Introduction

β-Blockers are the mainstay of modern therapy for heart failure. They reduce both mortality and morbidity and reverse multiple cellular and subcellular disturbances occurring in heart failure.1 However, the exact mechanism of their beneficial effects remains largely unknown. One of the postulated mechanisms is heart rate reduction (HRR), which improves left ventricular (LV) filling and coronary perfusion because of lengthened diastolic perfusion time, reduces myocardial O2 consumption, and improves contractility via reversed force–frequency relation. All of these effects are potentially beneficial. Indeed, pure heart-rate (HR)-reducing agents have been shown to improve many aspects of experimental heart failure.24

In humans, HR after an acute myocardial infarction (MI) is correlated with mortality5 and the beneficial effect of β-blockers is related to a quantitative HRR6 supporting the hypothesis that it is an important mechanism of β-blocker action. However, β-blockers exert numerous effects besides HRR, including anti-apoptotic, haemodynamic effects, improvement of myocardial energetics, and intracellular calcium handling.1 Thus the aim of our study was to compare haemodynamic, structural, cellular, and subcellular effects of β-blocker, metoprolol, and pure HR-reducing agent, ivabradine, in the rat model of the post-MI heart failure.

Both HR-reducing agents (ivabradine, zatebradine, anilidine)24,7 and β-blockers (metoprolol, bisoprolol, atenolol, carvedilol)811 have already been studied in the post-MI rat hearts. However, our study substantially extends those observations; (1) this is the first study to directly compare a pure HRR agent with a β-blocker, offering the same HRR, and (2) the first study to examine the effect of ivabradine on calcium handling in the post-MI rat heart; (3) our investigations were performed after washout of the drugs to reflect long-term changes instead of acute drugs effects; (4) drug administration was started 24 h after MI to avoid early antiarrhythmic effects, which could have introduced selection bias in some earlier studies.2,9

2. Methods

The investigation conformed to the Guide for the Care and Use of Laboratory Animals published by the US National Institutes of Health (NIH publication No. 85–23, revised 1996).

2.1 Animals and treatments

One hundred forty-four 10-week-old male Wistar rats (260–310 g) were used in the study. MI was induced in 120 rats, as described previously.12 Briefly, the rats were anaesthetized with ketamine HCl (100 mg/kg body weight, i.p.) and xylazine (5 mg/kg body weight, i.p.), and left thoracotomy was performed. The heart was externalized and a suture (5–0 silk) was placed around the proximal left coronary artery and tightly tied. The heart was internalized, the chest was closed, and pneumothorax was reduced. Twenty-four rats were subjected to the same protocol except that the snare was not tied. These rats served as the sham group. Within 24 h post-MI mortality was 25%.

Twenty-four hours after the surgery all animals underwent echocardiography and those with large MI (wall motion index, WMI < 16, n = 66) were randomly assigned to one of the three treatment groups: (1) receiving 10 mg/kg body weight/day of ivabradine (Servier, Courbevoie, France) in drinking water (n = 22); (2) receiving 250 mg/kg body weight/day of metoprolol (Polpharma, Poland) in drinking water (n = 22); and (3) no treatment (n = 22). The sham-operated rats were allocated to the same treatment groups. Forty-eight hours before the final examination at 8 weeks all drugs were discontinued to enable their washout. Eight weeks after the surgery all animals underwent echocardiography and LV catheterization. Subsequently, the hearts were excised, examined in Langendorff mode, and digested for cellular studies.

2.2 Echocardiography

Echocardiography was performed using MyLab25 (Esaote, Italy) with 13 MHz linear array transducer. Each rat was examined at baseline, 24 h, 7 days, (8 weeks–2 days), and 8 weeks after the surgery. Under light anaesthesia (ketamine HCl and xylazine, 75 and 3.5 mg/kg body weight, i.p.) LV end-diastolic and end-systolic diameters, as well as wall thickness, were determined from the short-axis view at the midpapillary level. Regional LV wall motion abnormalities were quantitiated as described previously.12 Contractility of 12 wall segments visualized in the midpapillary short-axis view and 11 segments visualized in the long-axis view was graded as 1 (normal) or 0 (abnormal) and the total WMI was calculated. The normal hearts had WMI = 23. Our previous results revealed that WMI closely correlated with infarct size and that WMI = 15 corresponded to infarct size ∼40%.12 LV end-diastolic and end-systolic areas were planimetered from the parasternal long-axis view. LV ejection fraction (LVEF) was calculated as (LV diastolic area − LV systolic area)/LV diastolic area. LV wall stress σm = PRi/2h(1+h/2Ri) was estimated using an equation by Falsetti et al.:13 Meridional wall stress, where P is LV end-diastolic pressure (LVEDP), Ri is LV end-diastolic diameter, and h is wall thickness.

2.3 Haemodynamic and Langendorff measurements

Under light anaesthesia a micromanometer-tipped catheter (Millar Instruments) was advanced through the right carotid artery into the LV for recording of LV pressures and peak rate of rise and decline of LV pressure (dP/dtmax and dP/dtmin). After baseline measurements, 15 s intravenous infusions of isoproterenol were given (2, 4, 6 µg/kg body weight) to obtain HR response as reported previosuly.14 Thereafter the rats were given time to restore baseline parameters. Next, after heparinization (1000 IU/kg body weight, i.p.) and pentobarbital anaesthesia (50 mg/kg body weight, i.p.), the heart was excised and Langendorff perfused with Krebs–Henseleit buffer. A saline-filled latex balloon, connected to a pressure transducer (Hugo Sachs, Germany) for pressure measurement, was introduced to the LV and gradually inflated, enabling the evaluation of LV diastolic pressure–volume relationships. Lungs and hearts were weighed.

2.4 Cells isolation and superfusion

Cells were isolated as described previously.15 After perfusion of the heart with collagenase (Boehringer) and protease (Sigma) containing solution, right ventricle was separated and discarded. Myocytes of the healthy LV tissue were resuspended in Tyrode solution, placed in the superfusion chamber mounted on the stage of an inverted microscope (Nikon), and rapidly superfused with Tyrode solution containing 2 mM Ca2+ at 37°C. Length and width of 30 randomly chosen myocytes from each heart were measured.

2.5 Ca2+ transients and cells shortening

The LV myocytes were incubated for 15 min with 10 µM Indo-1 acetoxymethyl ester as described by Spurgeon et al.16 The ratio of 405–495 nm Indo-1 fluorescence for the diastolic and systolic Ca2+ concentration was obtained from the output of Dual Channel Ratio Fluorometer (Biomedical Instrumentation Group, University of Pennsylvania). The difference between the systolic and diastolic Indo-1 ratios was used as a measure of the amplitude of Ca2+ transients. Cell shortening was recorded with video edge detector (Cardiovascular Laboratories, School of Medicine, UCLA).

2.6 Ca2+ transport by SERCA and NCX, sarcoplasmic reticulum Ca2+ content and diastolic SR Ca2+ leak

Ca2+ transport by sarcoplasmic reticulum (SR) Ca2+-ATPase (SERCA) and Na+/Ca2+ exchanger (NCX) was estimated from the rate constants (r) of single exponential curves fitted to decaying part of the electrically and caffeine-evoked Ca2+ transients, according to Choi and Eisner.17 Figure 1 presents the experimental protocol.

Figure 1

Experimental protocol used to assess Ca2+ transport by SERCA and NCX. Top panel: ventricular myocyte superfused with normal Tyrode solution (NT) was electrically paced at 1 Hz. Stimulation was stopped and 10 mM caffeine was rapidly superfused until complete relaxation of the caffeine induced Ca2+ transient. Bottom panel: the same myocyte again paced at 1 Hz and superfused with NT. Following 60 s of rest and perfusion with 0Na0Ca Tyrode solution, 10 mM caffeine was again applied. ISO2 Multitask-Patch-Clamp Software designed by M. Friedrich and K. Benndorf (University of Dresden) was used to fit monoexponetial curves to decaying part of Ca2+ transients and calculate r1, r2, and r3 rate constants. The rate of Ca2+ transport by SERCA and NCX was estimated according to formulas: rSERCA = r1 − r2 and rNCX = r2 − r3, respectively. Difference between diastolic Indo-1 fluorescence after 60s perfusion with 0Na0Ca solution with or without 1 mM tetracaine was taken as a measure of diastolic SR Ca2+ leak thought RyRs. a.u., arbitrary units.

The rate constant of decay of electrically evoked Ca2+ transients (r1) reflects the rate of combined Ca2+ transport by SERCA, which pumps Ca2+ back to SR, and by two sarcolemmal Ca2+ transporters: NCX and sarcolemmal Ca2+-ATPase (PMCA), which extrude Ca2+ from the cell (r1 = rSERCA+rNCX+rPMCA). Caffeine releases Ca2+ from the SR and prevents its reaccumulation (Ca2+ transport by SERCA under caffeine perfusion = 0). Thus, the rate constant of decline of caffeine-evoked Ca2+ transient (r2) reflects the rate of combined Ca2+ transport by NCX and PMCA (r2 = rNCX+rPMCA). When NCX is additionally blocked by perfusion with Na+,Ca2+-free solution (0Na0Ca—in Tyrode solution NaCl was replaced with LiCl and NaH2PO4 and CaCl2 were omitted), the rate constant of decline of caffeine-evoked Ca2+ transient (r3) reflects the rate of Ca2+ transport by PMCA (r3 = rPMCA). The rate constant of Ca2+ transport by SERCA was estimated by subtracting r2 from r1 (rSERCA = r1 − r2), while for estimation of the rate constant of Ca2+ transport by NCX, r3 was subtracted from r2 (rNCX = r2 − r3). Relative contribution to relaxation of the Ca2+ transporters was calculated according to formulas: SERCA contribution = r1 − r2/r1, NCX contribution = r2 − r3/r1.

To evaluate diastolic SR Ca2+ leak in some experiments tetracaine (1 mM), a blocker of ryanodine receptors (RyRs) was added to 0Na0Ca solution according to the method of Shannon et al.18 (Figure 1, lower panel). In resting myocytes superfused with 0Na0Ca solution, the main routes of Ca2+ influx and efflux are blocked. Therefore, the level of diastolic fluorescence is mostly dependent on spontaneous release of Ca2+ from the SR through RyRs. When tetracaine was added to 0Na0Ca solution, the resting fluorescence declined. The difference between diastolic fluorescence under 0Na0Ca with and without tetracaine perfusion was used as a measure of diastolic SR Ca2+ leak. Since diastolic Ca2+ leak is strictly dependent on SR Ca2+ load, to measure RyRs function (Ca2+ sensitivity) independently of SR Ca2+ content, the diastolic Ca2+ leak was normalized to SR Ca2+ content. SR Ca2+ content was estimated from the amplitude of caffeine-evoked Ca2+ transients in myocytes superfused with 0Na0Ca solution (Figure 1).

2.7 Statistical analysis

All results are given as means ± SEM. Differences for parameters obtained at each time point were evaluated by Student's t-test or by one-way ANOVA, followed by a Dunnett test in case of significance (Statistica software). Differences were considered significant at the level of P < 0.05.

3. Results

3.1 Mortality, infarct size, and heart rate

Eleven, eight, and 10 rats died in the untreated MI, MI + ivabradine, and MI + metoprolol groups, respectively, whereas no sham animal died between randomization (24 h after the surgery) and the study conclusion. Thus, the final sizes of the MI groups were 11, 14, and 12 animals, respectively. The mean infarct size in all MI groups, as assessed by echocardiographic WMI, was comparable both at randomization and at the study conclusion (Table 1).

View this table:
Table 1

Morphologic and echocardiographic parameters in the post-myocardial infarction and sham-operated rats

Untreated sham (n = 8)Untreated MI (n = 11)Sham + ivabradine (n = 8)MI + ivabradine (n = 14)Sham + metoprolol (n = 8)MI + metoprolol (n = 12)
Baseline BW, g284.5 ± 9.2278.4 ± 7.7273.3 ± 7.3274.3 ± 7.0283.0 ± 4.8283.8 ± 4.5
Final BW, g431.9 ± 15.9416.6 ± 16.1426.6 ± 22.0407.5 ± 20.1434.0 ± 16.8411.0 ± 16.3
Heart/BW, mg/g3.29 ± 0.104.77 ± 0.12*3.08 ± 0.105.50 ± 0.22*#3.13 ± 0.114.62 ± 0.17*$
LV weight/BW, mg/g2.48 ± 0.063.47 ± 0.08*2.32 ± 0.084.04 ± 0.17*#2.36 ± 0.083.40 ± 0.13*$
Lung weight/BW, mg/g6.22 ± 0.2813.63 ± 0.78*6.11 ± 0.3811.48 ± 0.52*#5.64 ± 0.5711.27 ± 0.77*#
Cardiomyocyte length, μm117.4 ± 2.8138.6 ± 3.4*120.8 ± 3.9136.5 ± 2.3*118.0 ± 2.8121 ± 2.3*#$
Cardiomyocyte width, μm22.9 ± 0.522.1 ± 0.622.1 ± 0.423.3 ± 0.521.9 ± 0.623.4 ± 0.5
Posterior wall end–diastolic thickness, mm1.35 ± 0.031.56 ± 0.03*1.38 ± 0.031.82 ± 0.04*#1.29 ± 0.021.55 ± 0.02*$
WMI at 24 h22.1 ± 0.113.3 ± 0.3*22.5 ± 0.113.2 ± 0.3*22.4 ± 0.113.2 ± 0.3*
WMI at 8 weeks22.0 ± 0.112.5 ± 0.4*22.1 ± 0.212.7 ± 0.5*22.3 ± 0.112.4 ± 0.4*
LVSP, mmHg129.0 ± 3.7104.5 ± 2.6*131.5 ± 3.9104.7 ± 3.7*133.0 ± 3.9105.7 ± 3.8*
LVEDP, mmHg6.4 ± 0.224.1 ± 0.5*5.7 ± 0.216.8 ± 1.4*#6.8 ± 0.316.7 ± 0.7*#
LVdevP, mmHg122.5 ± 1.680.3 ± 3.1*125.8 ± 1.887.8 ± 3.3*#136.6 ± 3.788.6 ± 2.8*#
+dP/dt, mmHg/s5744 ± 1153382 ± 120*5916 ± 1123849 ± 106*#5835 ± 1313785 ± 133*#
−dP/dt, mmHg/s4333 ± 902261 ± 93*4490 ± 1652792 ± 73*#4416 ± 2142537 ± 100*#$
LV diastolic wall stress, kdyn/cm27.9 ± 0.639.7 ± 1.0*7.5 ± 0.922.3 ± 2.4*#9.1 ± 0.926.2 ± 2.5*#
  • BW, body weight; MI, myocardial infarction; LV, left ventricle; WMI, wall motion index; LVSP, left ventricular systolic pressure; LVEDP, left ventricular end diastolic pressure; LVdevP, left ventricular developed pressure.

  • *P < 0.05 vs. corresponding sham.

  • #P < 0.05 vs. untreated MI.

  • $P < 0.05 vs. MI + ivabradine.

The baseline HR did not differ between all sham and MI animals (Figure 2). One week and (8 weeks–2 days) after the surgery, there were no differences in HRR induced by ivabradine and metoprolol, either in sham or in MI animals. Two days after ivabradine and metoprolol discontinuation (8 weeks after surgery) HR returned to similar values in all experimental groups. Peak HR in vivo (chronotropic response to 15 s infusion of isoproterenol) was reduced in MI animals. Ivabradine exacerbated this defect, whereas metoprolol tended to improve it (Figure 2B). As a result, chronotropic reserve (peak HR/HR at week 8), which was 88 ± 6% in sham animals, was reduced in MI and MI + ivabradine animals (37 ± 8% and 32 ± 4%, respectively), but was improved in MI + metoprolol animals (50 ± 5%) as compared with the MI and MI + ivabradine group (both P < 0.05). In sham animals intrinsic HR (iHR, HR measured in Langendorff perfused hearts) was by 28 ± 3% higher than HR recorded in vivo (P < 0.05; Figure 2A), whereas in MI and MI + ivabradine hearts was similar to that in vivo. Metoprolol partially prevented this chronotropic incompetence (Figure 2B).

Figure 2

Heart rate in the sham-operated (A) and post-myocardial infarction (B) rats, immediately before the surgery (baseline), 1 and 8 weeks–2 days (week 8–2 days) after the surgery, 2 days after discontinuation of ivabradine or metoprolol (week 8), peak response to isoproterenol infusion (ISO) and in Langendorff perfused hearts (iHR). Sham, sham-operated rats; MI, myocardial infarction. *P < 0.05 vs. untreated; #P < 0.05 vs. ivabradine; $P < 0.05 vs. corresponding sham. Black horizontal bar indicates the period of ivabradine or metoprolol administration. Number of experimental animals is provided in Table 1.

3.2 Haemodynamics, LV diameters, and morphological parameters

MI induced (1) LV systolic dysfunction, as evidenced by decreased LVEF (Figure 3C), LV peak systolic pressure (Table 1), and contractility (+dP/dt max, Table 1); (2) increase of LVEDP and pulmonary congestion, reflected by increased corrected lung weight (Table 1); (3) LV dilation, as evidenced by increased LV diastolic area (Figure 3B) and rightward shift of the diastolic pressure–volume curve (Figure 4); (4) LV hypertrophy, as indicated by increased corrected LV weight (Table 1), cardiomyocyte length (Table 1), and diastolic wall thickness (Table 1); and (5) almost five-fold increase of LV diastolic wall stress (Table 1). LV operating volume was increased by 120% (Figure 4).

Figure 3

Echocardiographic parameters in the sham-operated and post-myocardial infarction rats. Left ventricular systolic area (A), left ventricular diastolic area (B) and left ventricular ejection fraction (C). Sham, sham-operated rats; MI, myocardial infarction. *P < 0.05 vs. untreated MI. Number of experimental animals is provided in Table 1.

Figure 4

Left ventricular end diastolic pressure–volume relationships in Langendorff perfused hearts from the sham-operated and post-myocardial infarction rats. Operating volume, a parameter obtained by reading LV diastolic volume from diastolic pressure–volume curve obtained in Langendorff perfused heart, at LVEDP measured in vivo. LV, left ventricular; sham, sham-operated rats; MI, myocardial infarction; *P < 0.05 for operating volume vs. untreated MI. Number of experimental animals is provided in Table 1.

Ivabradine partially restored LV systolic function in the post-MI hearts, as evidenced by significantly higher LVEF (Figure 3C) and contractility (Table 1), lowered LVEDP, and reduced pulmonary congestion (Table 1). However, it did not prevent post-MI dilation, since LV diastolic area (Figure 3B) and operating volume (Figure 4) were unchanged. Moreover, there was even a trend towards rightward displacement of the LV diastolic pressure–volume curve (Figure 4). On the other hand, ivabradine potentiated post-MI hypertrophy. Although there was no further increase of cardiomyocyte dimensions above that observed in MI, both wall thickness and corrected LV weight were increased (Table 1). LV wall stress was reduced by 1.5-fold, as compared with untreated MI animals (Table 1).

Metoprotol restored LV function in the post-MI hearts to a similar degree as ivabradine, as evidenced by significantly higher LVEF (Figure 3C) and contractility (Table 1), as well as lowered LVEDP and reduced pulmonary congestion (Table 1). In contrast to ivabradine, it did partially prevent post-MI dilation (Figure 3B) and decreased LV operating volume (Figure 4), but had no effect on post-MI hypertrophy (cardiomyocytes were smaller but neither wall thickness nor corrected LV weight was different from that in MI). Metoprolol reduced LV wall stress by 1.5-fold, similarly as ivabradine, (Table 1). Although both ivabradine and metoprolol similarly partially restored contractility of the post-MI hearts (as evidenced by increased +dP/dt max), the maximal velocity of relaxation (−dP/dt max, Table 1) was improved more by ivabradine as compared with metoprolol (P < 0.05), indicating beneficial lusitropic effect of ivabradine.

3.3 Ca2+ handling

Analysis of variance did not show statistically significant differences with respect to Ca2+ transient amplitude and decay, SR Ca2+ content, Ca2+ transport by SERCA and NCX, and diastolic SR Ca2+ leak between the three sham groups (untreated sham, sham + ivabradine, sham + metoprolol). Thus, they were combined and treated as a single group (sham).

MI affected neither myocyte shortening (MI: 6.7 ± 0.7% vs. sham: 8.3 ± 0.5%, P > 0.05), nor the amplitude of Ca2+ transient (Figure 5B), nor SR Ca2+ content (Figure 5C). Rate of Ca2+ transient decay in the MI cardiomyocytes was decreased by 23.6% (Figure 5D). SERCA transporting function (rSERCA) in the MI cardiomyocytes was depressed by 35.3% (Figure 6A), whereas NCX function (rNCX) was increased by 58.3% (Figure 6B), thus the relative NCX contribution to relaxation increased from 6.2% in sham group to 12.6% in the MI group. Relative SERCA contribution to relaxation fell from 91.6% in sham group to 85.3% in the MI group (Figure 6C) and rNCX and rSERCA ratio in MI rose by as much as 116.7% as compared with sham (Figure 6D).

Figure 5

Parameters of Ca2+ transients and sarcoplasmic reticulum Ca2+ content after myocardial infarction—the effect of ivabradine and metoprolol. (A) Representative traces of Ca2+ transients in left ventricular myocytes. (B) Ca2+ transient amplitude, n = 62–134 myocytes. (C) Amplitude of caffeine-evoked Ca2+ transient during perfusion with 0Na 0Ca Tyrode solution (the index of sarcoplasmic reticulum Ca2+ content), n = 12–26. (D) Rate of Ca2+ transient decay (r1) in electrically stimulated myocytes (n = 12–16). Sham group is combined from the untreated sham, sham + ivabradine, sham + metoprolol groups. a.u., arbitrary units; sham, sham-operated rats; MI, myocardial infarction; Iva, ivabradine; Meto, metoprolol. *P < 0.05 vs. respective sham; #P < 0.05 vs. untreated MI; $P < 0.05 vs. MI + Iva.

Figure 6

Ca2+ transport by sarcoplasmic reticulum Ca2+-ATP-ase and Na+/Ca2+ exchanger after myocardial infarction—the effect of ivabradine and metoprolol. The mean rate constant of Ca2+ transport by sarcoplasmic reticulum Ca2+-ATP-ase (rSERCA) (A) and Na+/Ca2+ exchanger (rNCX) (B). (C) Relative contribution of SERCA and NCX to relaxation. (D) Ratio of the rate constant of Ca2+ transport by NCX and SERCA. n = 12–16 myocytes isolated from at least four animals. Sham group is combined from the untreated sham, sham + ivabradine, sham + metoprolol groups. Sham, sham-operated rats; MI, myocardial infarction; Iva, ivabradine; Meto, metoprolol. *P < 0.05 vs. respective sham; #P < 0.05 vs. untreated MI; $P < 0.05 vs. MI + Iva.

In the MI cardiomyocytes diastolic SR Ca2+ leak (Figure 7A) was increased by 67.9%. Ca2+ sensitivity of RyRs, assessed by normalization of diastolic SR Ca2+ leak to SR Ca2+ content, was increased by 47.8% as compared with the sham cardiomyocytes (Figure 7B). Since the SR Ca2+ content did not differ significantly between the sham and MI cardiomyocytes (Figure 5C), increased diastolic SR Ca2+ leak after MI can be predominantly attributed to the increased Ca2+ sensitivity of RyRs.

Figure 7

Diastolic sarcoplasmic reticulum Ca2+ leak after myocardial infarction—the effect of ivabradine and metoprolol. The mean diastolic sarcoplasmic reticulum (SR) Ca2+ leak (A) was calculated from the shift downward in resting Indo-1 fluorescence after 1 mM tetracaine addition to 0Na0Ca solution (Figure 1). To evaluate RyRs function independent on SR Ca2+ load, diastolic SR Ca2+ leak was normalized to SR Ca2+ content (B). n = 16–33 myocytes from at least three animals in each group. Sham group is combined from the untreated sham, sham + ivabradine, sham + metoprolol groups. Sham, sham-operated rats; MI, myocardial infarction; Iva, ivabradine; Meto, metoprolol. *P < 0.05 vs. respective sham; #P < 0.05 vs. untreated MI; $P < 0.05 vs. MI + Iva.

The Ca2+ transient amplitude was increased by 25.4% in the MI + ivabradine cardiomyocytes as compared with MI (Figure 5B), despite no change in SR Ca2+ content (Figure 5C). The mean rSERCA was improved by 19.3% (Figure 6A), whereas rNCX was unchanged as compared with the MI group, and highly increased as compared with sham (Figure 6B). Ivabradine normalized rate of Ca2+ transient decay, which was no longer different from that in sham group (Figure 5D). Relative contribution of the Ca2+ transporters to myocyte relaxation was unchanged as compared with MI (Figure 6C). The mean rNCX/rSERCA was unchanged as compared with the MI group markedly higher than in the sham group (Figure 6D). Both diastolic SR Ca2+ leak and normalized diastolic Ca2+ leak did not differ from those measured in the MI cardiomyocytes and were higher (even more distinctly than in the MI cardiomyocytes) than in the sham group (Figure 7A and B).

Metoprolol increased mean Ca2+ transient amplitude and SR Ca2+ content in the post-MI cardiomyocytes by 92.3 and 87.9%, respectively, as compared with the MI group. Unlike ivabradine, it had no effect on SERCA-transporting function (Figure 6A), but strongly depressed NCX-transporting function even below (by 30%) that recorded in the sham group (Figure 6B). These changes restored both rNCX/rSERCA and relative contribution of the Ca2+ transporters to relaxation in the MI + metoprolol group to values measured in the sham group (Figure 6C and D). However, Ca2+ transient decay, unlike in the MI + ivabradine group, was still depressed as compared with the sham group (Figure 5D). The highest diastolic SR Ca2+ leak was recorded in the MI + metoprolol groups. However, normalized Ca2+ leak in this group was no longer different from that in the sham group, was significantly lower as compared with MI + ivabradine group, and tended to be lower than in MI group.

4. Discussion

We found that ivabradine, a pure HRR agent, and metoprolol, a β-blocker, similarly effectively improved LV systolic function and reduced diastolic wall stress in post-MI heart failure in rats. However, mechanisms behind this protection were different. Metoprolol reduced LV dilation, which in combination with unchanged wall thickness reduced wall stress, whereas ivabradine did not prevent post-MI dilation but through potentiation of hypertrophy increased LV wall thickness, thus reducing wall stress. Furthermore, metoprolol markedly increased SR Ca2+ content and amplitude of Ca2+ transient, whereas ivabradine only mildly increased Ca2+ transient. Ivabradine, but not metoprolol, partially prevented MI-induced depression of SERCA function. Metoprolol, but not ivabradine, restored NCX overactivity after MI and normalized Ca2+ sensitivity of RyRs. Thus, although both drugs offered identical haemodynamic benefits in post-MI setting, metoprolol additionally prevented LV dilation, potentially deleterious hypertrophy, and chronotropic incompetence. Moreover, it potently increased cardiomyocyte contractility and had potentially antiarrhythmic cellular effects. This indicates that a pure HRR does not account for the effects of β-blockade on haemodynamic function, LV remodeling, and intracellular calcium handling in the post-MI setting.

In our study both ivabradine and metoprolol similarly reduced HR (by approximately 18%). The post-MI rats demonstrated chronotropic incompetence since both iHR and chronotropic reserve were decreased. Metoprolol, but not ivabradine, partially prevented this phenomenon. Since isolated hearts demonstrated chronotropic incompetence, it must be related to some structural changes rather than functional neurohumoral effects. Because chronotropic incompetence is associated with poor prognosis in human heart failure,19 its prevention by β-blocker may be of potential clinical importance.

4.1 Mechanism of beneficial effects of HRR and β-blocker on systolic function

All pure HRR agents have been reported to improve systolic function in the post-MI rat heart.24,7 Our results confirm these findings. Several mechanisms may participate in this phenomenon. First, ivabradine stimulated myocardial hypertrophy. Increased wall thickness on the one hand facilitates systolic performance, on the other hand reduces diastolic wall stress, an important driver of LV dilation. Our results contrast with the results of the study by Mulder et al.4 who did not find pro-hypertrophic effects of ivabradine in post-MI rat hearts. However, in their study ivabradine administration was started 7 days after MI, when major part of post-MI remodelling has already occurred,12 whereas in our study ivabradine was started 24 h after MI. On the other hand, Lei et al.3 have shown that administration of an HRR agent anilidine, started 24 h after MI in the rat, increased vascular endothelial growth factor (VEGF) expression in the heart, which was accompanied by significant LV hypertrophy. Second, HRR agents have been shown to improve coronary perfusion,3,7 by increasing diastolic time, potentiating angiogenesis, and exerting antifibrogenic effects,3 which could act to prevent deterioration of LV function. In our study ivabradine did not reduce LV dilation. Presumably HRR forced the heart to operate at the increased stroke volume and thus acted to facilitate LV dilation.

Metoprolol improved systolic LV function in the post-MI rat as shown previously,10,11 but in contrast to ivabradine, reduced LV dilation, and had no effect on diastolic wall thickness. There are several explanations for this: (1) β1-Receptors are coupled with multiple intracellular signalling pathways, exerting e.g. pro-apoptotic and pro-hypertrophic effects; their blockade could hypothetically prevent cardiomyocyte apoptosis and hypertrophy. (2) Activation of β1-receptors stimulates rennin–angiotensin system, which is also pro-apoptotic and pro-hypertrophic. (3) Unlike ivabradine, β-blockers have haemodynamic effects. They reduce afterload by decreasing blood pressure. This could result in reduced mechanical stimulus for both LV dilation and hypertrophy. On the other hand, reduced oxygen demand could act to diminish angiogenesis, which could additionally limit hypertrophy.7,11 (4) β-Blockers have been shown to improve calcium handling.

4.2 Ca2+ handling

Previous studies indicate that the pattern of calcium handling disturbances in post-MI rat cardiomyocytes strongly depends on experimental model, e.g. infarct size, time after MI, and study protocol: temperature, stimulation type, pacing frequency, and extracellular Ca2+ concentration.20,21 In our study cardiomyocytes were paced by electrical field and thus with their own potential (which is known to be prolonged after MI22), and extracellular Ca2+ concentration was 2 mM. These conditions were close to physiological. Pacing frequency was 1 Hz, which is well below physiological frequency (4–6 Hz), but this was necessary to obtain complete Ca2+ transient decay, which allowed for reliable rate constant (r1) calculation (see Methods and Figure 1).

4.2.1 SERCA- and NCX-transporting function

We used the rate of Ca2+ removal from cardiomyocyte cytoplasm as a measure of SERCA- and NCX-transporting function (see Methods). We feel that this parameter is the best indicator of protein function, since it reflects changes in the protein expression, functional regulation, and ionic environment influencing the transporting function. We found that SERCA function was decreased, whereas NCX function was increased in the post-MI rat cardiomyocytes. Decreased SERCA protein expression and/or mRNA levels have been observed in other studies in heart failure.10,23 Although β-blockers increased SERCA expression in some heart failure models,10 including humans,24 metoprolol did not improve SERCA function in our study. Ivabradine improved SERCA function, which likely resulted in greater improvement of relaxation in the MI + ivabradine group vs. MI + metoprolol group. Mechanism of this improvement is not clear. Increased NCX expression and/or function has been demonstrated in many heart failure models,25,26 including the post-MI rat heart.27 However, decreased expression and function have also been reported.28,29 We showed that although ivabradine had no effect on post-MI NCX function, metoprolol not only prevented MI-induced NCX overactivity, but also decreased its function below that observed in the sham group. This might have been related to the blockade of p38 kinase activation and resulting downregulation of NCX mRNA.30

Increased ratio of NCX- and SERCA-transporting function (rNCX/rSERCA) in post-MI cardiomyocytes leads to increased relative contribution of NCX to relaxation, which means that increased proportion of intracellular Ca2+ is removed from the cytoplasm by NCX as compared with SERCA. This could decrease cardiomyocte contractility and increase the inward current generated by NCX working in the forward mode, resulting in life-threatening arrhythmias.25 In our study metoprolol prevented this potentially deleterious increase of NCX-transporting function, whereas ivabradine did not.

4.2.2 Amplitude of Ca2+ transient and SR Ca2+ content and diastolic SR Ca2+ leak

Despite increased outward Ca2+ transport by NCX and decreased Ca2+ transport to SR, the amplitude of Ca2+ transient, SR Ca2+ content, and cardiomyocyte shortening were unchanged after MI. Others reported similar results,27,31 although decrease of these parameters was also observed.32 These differences probably reflect different experimental conditions and models used in these studies. Unchanged amplitude of Ca2+ transient and cell shortening could hypothetically result from prolonged action potential and resting depolarization, observed in the post-MI rats,33 promoting increased Ca2+ influx through NCX working in the reverse mode. Furthermore, prolonged action potential may increase Ca2+ influx through L-type calcium channels.34

The diastolic SR Ca2+ leak was enhanced after MI. Spontaneous diastolic SR Ca2+ release through RyRs depends on SR Ca2+ load and functional properties (Ca2+ sensitivity) of RyRs, which is related to the level of their PKA- and CaMKII-dependent phosphorylation. To obtain a measure of Ca2+ sensitivity of RyRs independent of SR Ca2+ load, we normalized diastolic Ca2+ leak to SR Ca2+ content, and we found that Ca2+ sensitivity of RyRs was increased after MI. Ivabradine tended to increase both the diastolic Ca2+ leak and Ca2+ sensitivity of RyRs. Increased Ca2+ sensitivity of RyRs may lead to increased fractional SR Ca2+ release and may explain increased amplitude of Ca2+ transient after ivabradine treatment, despite unchanged SR Ca2+ content in this group. We found the highest diastolic SR Ca2+ leak in the MI + metoprolol group. However, in this group SR Ca2+ content was also the highest among all experimental groups. Normalization of the diastolic Ca2+ leak to SR Ca2+ content revealed normal Ca2+ sensitivity of RyRs in MI + metoprolol group, as compared with that observed in the sham group. It is conceivable that metoprolol decreases PKA-dependent phosphorylation of RyRs35 and hereby protects RyRs functional properties after MI. Increased diastolic SR Ca2+ leak in the MI + metoprolol group did not lead to decreased SR Ca2+ content and amplitude of Ca2+ transient probably because of simultaneous decrease of NCX function (even below that recorded in sham group). Thus, spontaneously released Ca2+ from SR was preferably reuptaken by SERCA instead of being extruded from the cell by NCX.

4.3 Clinical implications

We show that although metoprolol and ivabradine, in doses producing a similar HRR in the post-MI rats, result in comparable preservation of LV systolic function, there are several concerns related to ivabradine. Ivabradine (1) did not prevent LV dilation, which was known to correlate with worse prognosis; (2) enhanced hypertrophy, an established risk factor of cardiovascular death; (3) did not prevent chronotropic incompetence, another marker of worse prognosis in heart failure; and (4) did not suppress potentially proarrhythmic NCX overactivity. However, its potential advantage over metoprolol is facilitation of relaxation. These concerns will need to be addressed in clinical trials, one of which is ongoing (BEAUTIFUL), before ivabradine is introduced to the field of the human heart failure therapy.

Conflict of interest: none declared.


This study was supported by a grant from Polish Cardiological Society.


  • Both authors equally contributed to the study.


View Abstract